Abstract
An important characteristic of many simulation experiments is the researcher's ability to induce correlations among the outputs or responses sampled during the study. Several strategies have been proposed for exploiting this opportunity to obtain correlated data yielding more reliable statistical inferences. The relative merit of these proposals becomes apparent when correlation induction is viewed as an integral part of the design of a simulation experiment. Simulations usually generate synthetic stochastic processes through the transformations of random number streams, that is, the sequences of uniform random variate between zero and one. The most important characteristic of these random number streams, other than their statistical validity, is that they be reproducible.
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have